Our AI Takes: The Limits of the Tools are Human

What happens when the limits of AI tools are not really technical — but human?

In this episode of Three Takes on AI, Brian, Campbell, and Mike talk through the AI tools they are actually using, where those tools are helping, and where they still fall short. From ChatGPT, Claude, Gemini, and Perplexity to image generation, workflow automation, coding, research, and productivity, the conversation gets practical fast.

But this episode is really about judgment.

AI can summarize, generate, search, automate, and accelerate, but it can also hallucinate, mislead, overpromise, and create false confidence. The real question is not just which tool is best. It is whether the humans using the tools understand their limits.

Campbell brings the governance lens.

Brian shares hands-on examples from building and experimenting with AI tools.

Mike grounds the discussion with a perfect real-world reminder: sometimes a human expert can see what AI completely misses.

This is a conversation about productivity, responsibility, small business adoption, governance, and why human experience still matters in an AI-driven world.